[noelpolar]

Are the photographers bald?

[3by2]

Thanks ! I hope other skeptikals can, at least, verify what we meant. Unfortunately I tried to zoom until 1600% with LR but I was not able to get any pattern: I just see purple pixels

BTW thanks mate!

Danilo

You mean a pattern like this. I can often see this pattern if I zoom in at 400%, not just in purple flare situations. I presume that's the sensor grid or some such technical thing. I can see it to a lesser extent on the K5 16Mp sensor sometimes (bigger grid). I guess if you're pushing the sensor you're going to see strange artefacts, similar to doing the same thing with film.

The more interesting question for me is whether the KP sensor is more sensitive to the effects of purple fringing/flare. Certainly I've got more examples with the KP than the K5 but I'm also shooting slightly different subjects in different ways with the KP than the K5 so actually having more examples proves nothing, I'd have to shoot them side by side in the same conditions.

[mctaveck]

You mean a pattern like this. I can often see this pattern if I zoom in at 400%, not just in purple flare situations. I presume that's the sensor grid or some such technical thing. I can see it to a lesser extent on the K5 16Mp sensor sometimes (bigger grid). I guess if you're pushing the sensor you're going to see strange artefacts, similar to doing the same thing with film.

Not exactly, the request is more related to the sensor pattern in terms of color pixel not the grid. If you go thru all the posts you can see one that is a link to an external source with phisycal explanation of the root cause. I cannot see , in my images any strange pattern or black or green pixels.

Thanks
Ciao

[house]

You mean a pattern like this. I can often see this pattern if I zoom in at 400%, not just in purple flare situations. I presume that's the sensor grid or some such technical thing. I can see it to a lesser extent on the K5 16Mp sensor sometimes (bigger grid). I guess if you're pushing the sensor you're going to see strange artefacts, similar to doing the same thing with film.

The more interesting question for me is whether the KP sensor is more sensitive to the effects of purple fringing/flare. Certainly I've got more examples with the KP than the K5 but I'm also shooting slightly different subjects in different ways with the KP than the K5 so actually having more examples proves nothing, I'd have to shoot them side by side in the same conditions.

That *is* the flare induced cross talk pattern I was referring to. It will look different on a xtrans camera and pdaf pixels may also influence the look.

The software used to process the files (including sooc) and what settings you use will influence the visibility of the pattern.

[RICHARD L.]

Could the creation of those artefacts depend on the lens you use ? I never see anything like that in my pictures. Here with an HD DFA 28-105 mm @ 28 mm FL and f/11.

[house]

As far as I understand the lens has to be wide or you have to provoke flare. I have also heard that ff lenses on crop bodies are more prone as light ends up where it shouldn't be.

Flare means light is bouncing around inside the lens hitting the sensor at angles it wasnt designed to handle.

So yes the lens is a big part of the equation. Presumably sensor stacks micro lenses etc also impact the performance.

[RICHARD L.]

Those artefacts can be created when using poorly coated filters, maybe. Modern SMC Pentax K lenses as well as 645 and 67 lenses don't flare at all. Straight lenses prove to be very resistant to flare.

DA 12-24 mm f/4

FA 28 mm f/2.8

P67 105 mm f/2.4

P67 105 mm f/2.4


645 200 mm f/4 (left) and P67 75 mm f/2.8 AL (right).

---

Last edited by RICHARD L.; 04-14-2021 at 02:25 PM.

[stevebrot]

- To figure out when it happens to avoid it

That is easy...don't do the things in the description on how to reproduce.

- Technical curiosity

Fair enough, though my bet rides on the lens/coatings side of things.

- Learn if it's a product failure to get your money back

Given that I have seen similar on film, I would say fat chance.


Steve

[stevebrot]

Just to muddy the waters a little, this magenta-themed thread from a year and a half ago might be interesting to some...

Shadow recovery issue on KP (magenta banding) - PentaxForums.com


Steve

[stevebrot]

You mean a pattern like this. I can often see this pattern if I zoom in at 400%, not just in purple flare situations.

Those are the pixels you have magnified 4x bigger. Whether you see the edges depends on the display software with many graphic editors showing them to allow pixel-level editing.

The more interesting question for me is whether the KP sensor is more sensitive to the effects of purple fringing/flare.

Same sensor as the K-3/K-3ii.


Steve

[house]

Those are the pixels you have magnified 4x bigger. Whether you see the edges depends on the display software with many graphic editors showing them to allow pixel-level editing.

Sorry you are mistaken. Or rather, it is of course the pixels, but the subject doesn't have that texture. There should be no repeating structure at pixel level like that. The pattern is in itself a minor issue but it's a symptom of cross talk between sensor pixels which ruin the image where it happens. Again just search the internet to find how Fuji users have reacted to the issue.

By tuning the demosaicing algorithm (if your software allows it) you can get rid of the pattern but not the purple flare.

[stevebrot]

Sorry you are mistaken.

I stand by what I wrote...uncorrected. FWIW, I am familiar with the X-Trans issues.


Steve

[3by2]

Those are the pixels you have magnified 4x bigger. Whether you see the edges depends on the display software with many graphic editors showing them to allow pixel-level editing.



Same sensor as the K-3/K-3ii.


Steve

Like I said, sometimes you can see that pattern, sometimes you can't, the magnification and the software are not the only factor. I have a couple of images where the patterns is obvious in one area and not the other.

Quite why that is I don't know, it seems quite similar to the Fuji manifestation.

Like I said I don't just see it in flare situations, higher iso, lower light also produces it but again, not everywhere.

[stevebrot]

If you go thru all the posts you can see one that is a link to an external source with phisycal explanation of the root cause.

Are you referring to a link on this thread or on FB. If on FB, it would be amazingly helpful to simply provide the link. Even searching the group for "campanelli post" failed to bring up a meaningful list of posts.

Steve

[stevebrot]

My conclusions:

• Since several examples have come from cameras other than the KP, this appears to be other than a KP phenomenon.
• Without access to the DNG/PEF file for a capture showing the effect, this matter is not worth pursuing further and no conclusions are possible since processing artifact cannot be ruled out.

Adios!


Steve